The first cases of pediatric AIDS in Miami were born in 1979. The first diagnoses were made in 1982 and through September, 1985, 74 cases of HTLV-III infection in infants and children have been identified in 60 households. The initial enrollment period was from March, 1983 until October, 1984. Seventeen index cases and an equal number of matched comparison cases and their respective households were enrolled and followed longitudinally since enrollment. The follow-up rate on this group exceeded 90% in the current grant period. Diagnostic and laboratory parameters including HTLV-III isolation and serologic procedures were established and pediatric AIDS has been defined in this group. The mothers of perinatally-infected index cases are themselves infected with HTLV-III and have profound immune dysfunction and although clinically healthy at the time of delivering an infected infant, a significant number of the mothers subsequently develop AIDS. The mothers are persistently infected with the virus and may give birth to multiply affected sibships. Of 20 infants born to mothers after a previously identified pediatric AIDS index case, 65% of the subsequent children are infected. In contrast to this high rate of perinatal infection, there has been no evidence of casual infection noted in households. The overall case fatality rate is 35%, although in certain clinical syndromes such as lymphocytic interstitial pneumonitis, the rate is only 14% suggesting that there may be distinct clinical outcomes perhaps associated with different immune responses to the virus. It is proposed to enroll all surviving index cases diagnosed before and after the initial enrollment period and to follow a total of some 75 pediatric AIDS cases for four more years to enlarge our data base and to extend the period of follow-up. This will provide information on the late outcome of pediatric AIDS, the possibility of HTLV-III transmission in households and an evaluation of late outcome in HTLV-III virus-positive mothers. Virus isolates will be characterized and antibody responses to HTLV-III will be measured to evaluate the respective roles of viral variation, antibody formation and clinical disease.